A Probabilistic Propagation Graph Modeling Method for Channel Characterization in Dense Urban Environments

Abstract

With the rapid development of mobile broadband communications such as the “fifth generation (5G)” systems, accurate and efficient simulation-based channel modeling becomes more essential for propagation channel characterization in the complicated environments. In the previous works, the so-called “propagation-graph-based channel modeling” method has been introduced and effectively applied to the simulation of channel characteristics. Compared with the conventional ray-tracing-based approaches, the graph-based modeling may significantly reduce the computational complexity. In this contribution, without increasing computational complexity, the graph-based channel modeling method is proposed for channel characterization in the dense urban environments with the following improvements, i.e. realistic geometry-assisted scattering point placement, scattering coefficient definition, and line-of-sight (LoS) path presence probability settings. By comparing the simulated channel characteristics with those obtained in the real measurements in two urban scenarios, the proposed improvements are demonstrated in predicting channel characteristics more accurately.